Fix incorrect problem IDs

README.md

@@ -36,7 +36,7 @@ The PEtab files are available in the respective subdirectories.
 | [Fiedler_BMCSystBiol2016](problems/Fiedler_BMCSystBiol2016/)                         | [v1](problems/Fiedler_BMCSystBiol2016/v1/)             |                                           | Reaction network model of RAF/MEK/ERK signaling with ERK feedback in HeLa cells, fitted to phospho-MEK/ERK data without and with RAF or MEK inhibitors.          |
 | [Froehlich_CellSystems2018](problems/Froehlich_CellSystems2018/)                     | [v1](problems/Froehlich_CellSystems2018/v1/)           |                                           | Large-scale reaction network model of cancer signaling pathways, parameterized by cell-line-specific gene expression and fitted to drug response viability data. |
 | [Fujita_SciSignal2010](problems/Fujita_SciSignal2010/)                               | [v1](problems/Fujita_SciSignal2010/v1/)                |                                           | Reaction network model of EGF-induced EGFR, Akt and S6 phosphorylation in PC12 cells under different EGF stimulation profiles.                                   |
-| [Giordano_Nature2020](problems/Giordano_Nature2020/)                                 | [v1](problems/Giordano_Nature2020/v1/)                 |                                           | Compartmental epidemiological model (SIDARTHE, distinguishing detected/undetected and symptom severity) of the COVID-19 epidemic in Italy.                       |
+| [Giordano_NatMed2020](problems/Giordano_NatMed2020/)                                 | [v1](problems/Giordano_NatMed2020/v1/)                 |                                           | Compartmental epidemiological model (SIDARTHE, distinguishing detected/undetected and symptom severity) of the COVID-19 epidemic in Italy.                       |
 | [Isensee_JCB2018](problems/Isensee_JCB2018/)                                         | [v1](problems/Isensee_JCB2018/v1/)                     |                                           | Reaction network model of cAMP-induced PKA activation and phosphorylation of its regulatory subunit in rat sensory neurons under pharmacological perturbations.  |
 | [Lang_PLOSComputBiol2024](problems/Lang_PLOSComputBiol2024/)                         | [v1](problems/Lang_PLOSComputBiol2024/v1/)             |                                           | Reaction network model of the human cell cycle, generated from a rule-based model and fitted to compartment-resolved single-cell imaging data of RPE-1 cells.    |
 | [Laske_PLOSComputBiol2019](problems/Laske_PLOSComputBiol2019/)                       | [v1](problems/Laske_PLOSComputBiol2019/v1/)            |                                           | Reaction network model of intracellular influenza A virus replication, fitted to viral RNA, genome import and virus release data in human A549 cells.            |
@@ -49,10 +49,10 @@ The PEtab files are available in the respective subdirectories.
 | [Raia_CancerResearch2011](problems/Raia_CancerResearch2011/)                         | [v1](problems/Raia_CancerResearch2011/v1/)             |                                           | Reaction network model of IL-13-induced JAK2/STAT5 signaling in the primary mediastinal B-cell lymphoma cell line MedB-1.                                        |
 | [Raimundez_PCB2020](problems/Raimundez_PCB2020/)                                     | [v1](problems/Raimundez_PCB2020/v1/)                   |                                           | Reaction network model of EGFR and MET signaling to ERK and AKT in two gastric cancer cell lines in response to EGF and the anti-EGFR antibody cetuximab.        |
 | [SalazarCavazos_MBoC2020](problems/SalazarCavazos_MBoC2020/)                         | [v1](problems/SalazarCavazos_MBoC2020/v1/)             |                                           | Reaction network model of EGFR dimerization, multisite phosphorylation and binding of the adaptor proteins GRB2 and SHC1, generated from a rule-based model.     |
-| [Schwen_PONE2014](problems/Schwen_PONE2014/)                                         | [v1](problems/Schwen_PONE2014/v1/)                     | [v2](problems/Schwen_PONE2014/v2/)        | Reaction network model of insulin binding, internalization and degradation in two subpopulations of isolated hepatocytes with low and high insulin binding.      |
+| [Schwen_PLoSOne2015](problems/Schwen_PLoSOne2015/)                                   | [v1](problems/Schwen_PLoSOne2015/v1/)                  | [v2](problems/Schwen_PLoSOne2015/v2/)     | Reaction network model of insulin binding, internalization and degradation in two subpopulations of isolated hepatocytes with low and high insulin binding.      |
 | [Smith_BMCSystBiol2013](problems/Smith_BMCSystBiol2013/)                             | [v1](problems/Smith_BMCSystBiol2013/v1/)               |                                           | Large reaction network model of insulin signaling in rodent adipocytes and its regulation by oxidative stress.                                                   |
 | [Sneyd_PNAS2002](problems/Sneyd_PNAS2002/)                                           | [v1](problems/Sneyd_PNAS2002/v1/)                      |                                           | Kinetic model of the type-2 IP3 receptor, fitted to time courses of calcium release from liver microsomes at varying IP3 and calcium concentrations.             |
-| [Weber_BMC2015](problems/Weber_BMC2015/)                                             | [v1](problems/Weber_BMC2015/v1/)                       |                                           | Reaction network model of interactions of protein kinase D, a lipid kinase (PI4KIIIbeta) and a ceramide transfer protein (CERT) at the trans-Golgi network.      |
+| [Weber_BMCSystBiol2015](problems/Weber_BMCSystBiol2015/)                             | [v1](problems/Weber_BMCSystBiol2015/v1/)               |                                           | Reaction network model of interactions of protein kinase D, a lipid kinase (PI4KIIIbeta) and a ceramide transfer protein (CERT) at the trans-Golgi network.      |
 | [Zhao_QuantBiol2020](problems/Zhao_QuantBiol2020/)                                   | [v1](problems/Zhao_QuantBiol2020/v1/)                  |                                           | Compartmental epidemiological model of COVID-19 in Wuhan, Hubei and the rest of China, with quarantined and unquarantined infected individuals.                  |
 | [Zheng_PNAS2012](problems/Zheng_PNAS2012/)                                           | [v1](problems/Zheng_PNAS2012/v1/)                      |                                           | Reaction network model of the combinatorial methylation of lysines 27 and 36 on histone H3.                                                                      |
 <!-- END OVERVIEW TABLE -->

problems/Giordano_Nature2020/v1/model.xml → problems/Giordano_NatMed2020/v1/model.xml

@@ -1,7 +1,7 @@
 <?xml version='1.0' encoding='UTF-8'?>
 <!-- Created by COPASI version 4.29 (Build 228) on 2020-09-17 20:49 with libSBML version 5.18.0. -->
 <sbml xmlns="http://www.sbml.org/sbml/level2/version4" level="2" version="4">
-  <model metaid="COPASI0" id="Giordano_Nature2020" name="Giordano_Nature2020">
+  <model metaid="COPASI0" id="Giordano_NatMed2020" name="Giordano_NatMed2020">
     <notes>
       <body xmlns="http://www.w3.org/1999/xhtml">
         <pre>This is a model created on COPASI 4.27 (Build 217) which reproduces the Figures 2b, 2d, 3b, 3d, 4b, 4d in the article - https://www.nature.com/articles/s41591-020-0883-7

problems/Schwen_PONE2014/v1/model.xml → problems/Schwen_PLoSOne2015/v1/model.xml

@@ -1,6 +1,6 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <sbml xmlns="http://www.sbml.org/sbml/level2/version4" level="2" version="4">
-  <model id="Schwen_PONE2014" name="Schwen_PONE2014">
+  <model id="Schwen_PLoSOne2015" name="Schwen_PLoSOne2015">
     <notes>
       <body xmlns="http://www.w3.org/1999/xhtml">
         <pre>PEtab implementation of the model from Schwen et al. (2015), PLoS One, 10, e0133653</pre>

problems/Schwen_PONE2014/v2/model.xml → problems/Schwen_PLoSOne2015/v2/model.xml

@@ -1,6 +1,6 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <sbml xmlns="http://www.sbml.org/sbml/level2/version4" level="2" version="4">
-  <model id="Schwen_PONE2014" name="Schwen_PONE2014">
+  <model id="Schwen_PLoSOne2015" name="Schwen_PLoSOne2015">
     <notes>
       <body xmlns="http://www.w3.org/1999/xhtml">
         <pre>PEtab implementation of the model from Schwen et al. (2015), PLoS One, 10, e0133653</pre>

problems/Weber_BMC2015/v1/model.xml → problems/Weber_BMCSystBiol2015/v1/model.xml

@@ -1,6 +1,6 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <sbml xmlns="http://www.sbml.org/sbml/level2/version4" level="2" version="4">
-  <model id="Weber_BMC2015" name="Weber_BMC2015">
+  <model id="Weber_BMCSystBiol2015" name="Weber_BMCSystBiol2015">
     <notes>
       <body xmlns="http://www.w3.org/1999/xhtml">
         <pre>PEtab implementation of the model from Weber et al. (2015), BMC Syst. Biol., 9, 9</pre>

src/python/benchmark_models_petab/C.py

@@ -10,6 +10,14 @@
 MODELS: List[str] = sorted(os.listdir(MODELS_DIR))
 MODEL_DIRS: List[str] = [os.path.join(MODELS_DIR, d) for d in MODELS]
 
+#: Former problem IDs, mapped to the current ones.
+#:  Accepted by the accessor functions, but not part of `MODELS`.
+RENAMED_MODELS: dict[str, str] = {
+    "Giordano_Nature2020": "Giordano_NatMed2020",
+    "Schwen_PONE2014": "Schwen_PLoSOne2015",
+    "Weber_BMC2015": "Weber_BMCSystBiol2015",
+}
+
 # layout of a single problem directory (`<MODELS_DIR>/<ProblemID>/...`)
 V1_DIRNAME: str = "v1"
 V2_DIRNAME: str = "v2"

src/python/benchmark_models_petab/__init__.py

@@ -6,7 +6,7 @@
 
 from . import v1, v2
 from .v1 import get_problem, get_problem_yaml_path, get_simulation_df
-from .C import MODEL_DIRS, MODELS, MODELS_DIR
+from .C import MODEL_DIRS, MODELS, MODELS_DIR, RENAMED_MODELS
 from .metadata import ProblemMetadata, get_metadata
 from .overview import get_overview_df
 from importlib.metadata import PackageNotFoundError, version

src/python/benchmark_models_petab/_ids.py

@@ -0,0 +1,22 @@
+"""Handling of former problem IDs."""
+
+import warnings
+
+from .C import RENAMED_MODELS
+
+
+def resolve_id(id_: str) -> str:
+    """Map a former problem ID to the current one, with a warning.
+
+    Other IDs are returned unchanged.
+    """
+    if (new_id := RENAMED_MODELS.get(id_)) is None:
+        return id_
+
+    warnings.warn(
+        f"Problem `{id_}` was renamed to `{new_id}`. "
+        "The old ID will stop working in a future version.",
+        FutureWarning,
+        stacklevel=3,
+    )
+    return new_id

src/python/benchmark_models_petab/metadata.py

@@ -8,6 +8,7 @@
 import yaml
 from pydantic import BaseModel, ConfigDict, Field
 
+from ._ids import resolve_id
 from .C import METADATA_FILENAME, MODELS_DIR
 
 METADATA_SCHEMA_FILENAME = "metadata.schema.json"
@@ -66,7 +67,7 @@ def get_metadata_path(id_: str) -> Path:
     -------
     The path to the metadata file. The file is not guaranteed to exist.
     """
-    return Path(MODELS_DIR, id_, METADATA_FILENAME)
+    return Path(MODELS_DIR, resolve_id(id_), METADATA_FILENAME)
 
 
 def get_metadata(id_: str) -> ProblemMetadata:

src/python/benchmark_models_petab/v1.py

@@ -6,6 +6,7 @@
 
 import pandas as pd
 
+from ._ids import resolve_id
 from .C import (
     MODELS as ALL_MODELS,
     MODELS_DIR,
@@ -33,6 +34,7 @@ def get_problem_yaml_path(id_: str) -> Path:
     -------
     The path to the PEtab problem YAML file.
     """
+    id_ = resolve_id(id_)
     yaml_path = Path(MODELS_DIR, id_, V1_DIRNAME, PROBLEM_FILENAME)
     if not yaml_path.exists():
         raise ValueError(
@@ -69,6 +71,7 @@ def get_simulation_df(id_: str) -> pd.DataFrame | None:
     -------
     The simulation dataframe if it exists, else None.
     """
+    id_ = resolve_id(id_)
     path = Path(MODELS_DIR, id_, V1_DIRNAME, SIMULATIONS_FILENAME)
     if path.is_file():
         return petab.get_simulation_df(path)

src/python/benchmark_models_petab/v2.py

@@ -5,6 +5,7 @@
 import pandas as pd
 import petab.v2 as petab
 
+from ._ids import resolve_id
 from .C import (
     MODELS as ALL_MODELS,
     MODELS_DIR,
@@ -32,6 +33,7 @@ def get_problem_yaml_path(id_: str) -> Path:
     -------
     The path to the PEtab v2 problem YAML file.
     """
+    id_ = resolve_id(id_)
     yaml_path = Path(MODELS_DIR, id_, V2_DIRNAME, PROBLEM_FILENAME)
     if not yaml_path.exists():
         raise ValueError(
@@ -67,6 +69,7 @@ def get_simulation_df(id_: str) -> pd.DataFrame | None:
     -------
     The simulation dataframe if it exists, else None.
     """
+    id_ = resolve_id(id_)
     path = Path(MODELS_DIR, id_, V2_DIRNAME, SIMULATIONS_FILENAME)
     if path.is_file():
         return pd.read_csv(

src/python/test/test_renamed.py

@@ -0,0 +1,36 @@
+"""Tests for former problem IDs."""
+
+import pytest
+
+import benchmark_models_petab as models
+from benchmark_models_petab.overview import get_readme_table
+
+
+def test_renamed_models_are_not_problems():
+    """Former IDs are no separate problems, their replacements exist."""
+    assert not set(models.RENAMED_MODELS) & set(models.MODELS)
+    assert not set(models.RENAMED_MODELS) & set(models.v1.MODELS)
+    assert not set(models.RENAMED_MODELS) & set(models.v2.MODELS)
+    assert set(models.RENAMED_MODELS.values()) <= set(models.MODELS)
+
+    table = get_readme_table()
+    for old_id in models.RENAMED_MODELS:
+        assert old_id not in table
+
+
+@pytest.mark.parametrize(
+    "old_id,new_id",
+    [("Schwen_PONE2014", "Schwen_PLoSOne2015")],
+)
+def test_former_id_is_accepted_with_warning(old_id, new_id):
+    for version in (models.v1, models.v2):
+        with pytest.warns(FutureWarning, match=new_id):
+            path = version.get_problem_yaml_path(old_id)
+        assert path == version.get_problem_yaml_path(new_id)
+    with pytest.warns(FutureWarning, match=new_id):
+        problem = models.get_problem(old_id)
+    assert problem.model.model_id == new_id
+    with pytest.warns(FutureWarning, match=new_id):
+        assert models.get_simulation_df(old_id) is not None
+    with pytest.warns(FutureWarning, match=new_id):
+        assert models.get_metadata(old_id) == models.get_metadata(new_id)